Clinical research data safety equipment wearing state monitoring method and system based on artificial intelligence

Abstract

The invention relates to the technical field of brain disease monitoring, in particular to a clinical research data safety equipment wearing state monitoring method and system based on artificial intelligence. Comprising the following steps: S1, acquiring body information and a brain disease type of a patient, and acquiring a wearable equipment type at the same time; s2, acquiring brain physiological data of the patient in real time through the wearable equipment, monitoring state data at the same time of data acquisition of the wearable equipment, and performing data anomaly threshold setting on the brain physiological data according to the state data in combination with a brain disease type; the abnormal threshold value is dynamically adjusted through the difference between the real-time state data and the standard state of the wearable equipment, the worse the equipment state is, the larger the allowed physiological data fluctuation range is, false alarms caused by non-pathological factors such as poor equipment contact and insufficient electric quantity are avoided, the abnormal threshold value is automatically adjusted, and false positive data are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of brain disease monitoring. Specifically, it relates to a method and system for monitoring the wearing state of clinical research data security equipment based on artificial intelligence. Background Art

[0002] With the development of living standards and the improvement of medical standards, human health problems have received more and more extensive attention. For the elderly and sub-healthy populations, real-time collection of their physiological indicators can effectively monitor their health status, achieve the purpose of disease prevention and early treatment, and provide data support for disease diagnosis, efficacy evaluation and safety warning by continuously collecting brain physiological data and equipment status information;

[0003] Currently, when conducting health analysis, it generally works in a fixed parameter mode, that is, a unified abnormal threshold of physiological data is preset, and the real-time state of the wearable equipment is not included in the analysis dimension. In the EEG monitoring of epilepsy patients, only whether the amplitude of the electroencephalogram waveform exceeds the fixed threshold is used to judge abnormality, while ignoring signal interference caused by non-pathological factors such as poor electrode contact and insufficient battery power of the equipment. This will lead to a large number of false positive alarms caused by abnormal equipment status, increasing the ineffective workload of medical staff, and at the same time being unable to capture the time correlation of the data sequence, and may misdelete real abnormal data with clinical value. Therefore, a method and system for monitoring the wearing state of clinical research data security equipment based on artificial intelligence are proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for monitoring the wearing state of clinical research data security equipment based on artificial intelligence to solve the problems raised in the above background art.

[0005] To achieve the solution of the above technical problems, one of the purposes of the present invention is to provide a method for monitoring the wearing state of clinical research data security equipment based on artificial intelligence, including the following steps:

[0006] S1. Obtain the patient's physical information and the type of brain disease, and at the same time obtain the type of wearable equipment;

[0007] S2. Real-time obtain the patient's brain physiological data through the wearable equipment, and at the same time monitor the status data of the wearable equipment at the same time when collecting data, and set the data abnormal threshold for the brain physiological data according to the status data combined with the type of brain disease;

[0008] S3. Use the AI model to perform abnormal data analysis on the historical brain physiological data combined with the status data and the data abnormal threshold, and at the same time extract the development trend of the adjacent time periods of the abnormal data, and make normal adjustments to the abnormal data according to the development trend;

[0009] S4. The AI model collects relevant patient data of the same type of brain disease, then combines the collected relevant patient data with the adjusted historical brain physiological data to predict the development data of the brain disease, and generates corresponding predicted brain physiological data based on the predicted development data of the brain disease;

[0010] S5. Set accurate thresholds and effective ranges according to the disease status of the brain disease type, compare the real-time brain physiological data combined with the predicted brain physiological data with the accurate thresholds, determine the effective range of the predicted brain physiological data according to the comparison results, and give safety reminders to the patients based on the predicted brain physiological data.

[0011] As a further improvement of this technical solution, in S1, the hospital management terminal is used to obtain the patient's physical information, and at the same time, the physical information related to the type of brain disease is extracted from the physical information for storage;

[0012] The hospital management terminal is used to obtain the patient's medical examination report, extract the corresponding type of brain disease of the patient and the disease status of the type of brain disease from the medical examination report, and at the same time extract the type of wearable device equipped for the patient from the medical examination report.

[0013] As a further improvement of this technical solution, the wearable device types in S1 include electroencephalogram caps, magnetic resonance imaging monitoring devices for the brain, and smart headbands. By synchronizing the wearable device types with the hospital management terminal through a wireless protocol, the obtained brain physiological data is synchronized to the clinical system, providing a data basis for subsequent safety reminders.

[0014] As a further improvement of this technical solution, the steps of S2 are as follows:

[0015] S2.1. Set the acquisition frequency for the wearable device according to the patient's physical information and the type of brain disease, and then the wearable device regularly acquires the patient's brain physiological data according to the set acquisition frequency;

[0016] S2.2. Extract the status data of the wearable device, and at the same time extract the status data at the same time according to the time of collecting the brain physiological data;

[0017] S2.3. Obtain the standard status data of the patient's wearable device, and then set the data anomaly threshold for the brain physiological data by combining the standard status data and the extracted status data with the type of brain disease;

[0018] The greater the difference value between the extracted status data and the standard status data, the higher the data anomaly threshold;

[0019] The smaller the difference value between the extracted status data and the standard status data, the lower the data anomaly threshold. When the extracted status data is the same as the standard status data, the data anomaly threshold is 0.

[0020] As a further improvement of this technical solution, the steps of S3 are as follows:

[0021] S3.1. Establish an AI model based on a deep learning algorithm;

[0022] S3.2. Perform abnormal data analysis by combining historical brain physiological data with state data at the same time and the data abnormality threshold corresponding to each state data. When the difference between historical brain physiological data at a certain moment and historical brain physiological data at the previous moment exceeds the data abnormality threshold, it is determined that the historical brain physiology at that moment is abnormal data. Conversely, when the difference between historical brain physiological data at a certain moment and historical brain physiological data at the previous moment does not exceed the data abnormality threshold, it is determined that the historical brain physiology at that moment is normal data;

[0023] S3.3. Obtain the development trend of historical brain physiological data, and at the same time extract the development trend of adjacent time periods of abnormal data. Calibrate the abnormal data according to the development trend, and then make normal adjustments to the abnormal data after data calibration.

[0024] As a further improvement of this technical solution, during the process of data calibration of abnormal data in S3.3, monitor the condition status of brain diseases. When the condition is adjusted in the adjacent time period of the abnormal data acquisition moment, do not perform data calibration on the abnormal data, and feedback the abnormal data to the hospital management terminal for manual adjustment.

[0025] As a further improvement of this technical solution, the steps of S4 are as follows:

[0026] S4.1. Collect all patient data through the AI model at the hospital management terminal, extract relevant patient data according to different brain disease types from the patient data, and then only save the relevant patient data of the same brain disease type for this patient safety reminder;

[0027] S4.2. Combine the relevant patient data with the adjusted historical brain physiological data to predict the development data of brain diseases for the patients of this safety reminder, obtain the predicted development data of brain diseases corresponding to the patients, and then generate predicted brain physiological data according to the predicted development data of brain diseases.

[0028] As a further improvement of this technical solution, the steps of S5 are as follows:

[0029] S5.1. Set accurate thresholds and effective ranges according to the condition status of brain diseases;

[0030] S5.2. Combine the real-time brain physiological data with the predicted brain physiological data and compare them with the accurate threshold. When the comparison result shows that the deviation between the real-time brain physiological data and the predicted brain physiological data is less than the accurate threshold, extract the predicted brain physiological data within the corresponding range according to the effective range, and give a safety reminder to the patient based on the predicted brain physiological data within the corresponding range;

[0031] S5.3. When the comparison result shows that the deviation between the real-time brain physiological data and the predicted brain physiological data is greater than the accurate threshold, continue the monitoring.

[0032] The second object of the present invention is to provide a wearable state monitoring system for clinical research data security equipment based on artificial intelligence, including a wearable state monitoring method for clinical research data security equipment based on artificial intelligence as described in any one of the above, including a data management module, a data prediction module, and a prediction comparison module;

[0033] The data management module is used to obtain the patient's physical information and brain disease type, and at the same time obtain the wearable equipment type. At the same time, monitor the status data of the wearable equipment at the same moment when collecting data, and set the data anomaly threshold for the brain physiological data according to the status data combined with the brain disease type;

[0034] The data prediction module is used to use the AI model to analyze the abnormal data by combining the historical brain physiological data with the status data and the data anomaly threshold, and at the same time extract the development trend of the adjacent time periods of the abnormal data, and make normal adjustments to the abnormal data according to the development trend. The AI model collects the relevant patient data of the same brain disease type, and then combines the collected relevant patient data with the adjusted historical brain physiological data to predict the brain disease development data, and generates the corresponding predicted brain physiological data according to the predicted brain disease development data;

[0035] The prediction comparison module is used to set the accurate threshold and the effective range according to the disease status of the brain disease type, compare the real-time brain physiological data combined with the predicted brain physiological data with the accurate threshold, determine the effective range of the predicted brain physiological data according to the comparison result, and give a safety reminder to the patient based on the predicted brain physiological data.

[0036] Compared with the prior art, the beneficial effects of the present invention:

[0037] A method and system for monitoring the wearing state of clinical research data security equipment based on artificial intelligence dynamically adjusts the abnormal threshold according to the difference between the real-time state data of the wearable equipment and the standard state. The worse the equipment state is, the larger the allowable fluctuation range of physiological data is, avoiding false alarms caused by non-pathological factors such as poor equipment contact and insufficient power, automatically adjusting the abnormal threshold, reducing false positive data, and calibrating abnormal data deviating from the trend to ensure the coherence of the data sequence. If there is a condition adjustment in the adjacent period of abnormal data, the automatic calibration is skipped and transferred to manual processing to avoid the algorithm covering the real data of clinical intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is the overall flow block diagram of the present invention;

[0039] Figure 2 It is the flow block diagram of the present invention for extracting the state data at the same time according to the time of collecting the brain physiological data;

[0040] Figure 3 It is the flow block diagram of the present invention for establishing an AI model based on a deep learning algorithm;

[0041] Figure 4 It is the flow block diagram of the present invention for collecting all patient data through the AI model at the hospital management end;

[0042] Figure 5 It is the flow block diagram of the present invention for giving a safety reminder to the patient according to the predicted brain physiological data within the corresponding range. DETAILED DESCRIPTION OF THE INVENTION

[0043] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work shall fall within the protection scope of the present invention.

[0044] As Figure 1 - Figure 5 shown, one of the purposes of the present invention is to provide a method for monitoring the wearing state of clinical research data security equipment based on artificial intelligence, including the following steps:

[0045] S1. Obtain the patient's physical information and brain disease type, and at the same time obtain the type of wearable equipment;

[0046] The S1 obtains the patient's physical information through the hospital management end, and at the same time extracts and saves the physical information related to the brain disease type from the physical information;

[0047] Obtain the patient's medical examination report through the hospital management terminal, extract the corresponding brain disease type of the patient and the condition status of the brain disease type from the medical examination report, and at the same time extract the type of wearable device equipped for the patient from the medical examination report;

[0048] Obtain the patient's medical report from the hospital management system, and according to the diagnostic information in the report, extract the patient's brain disease type. For example, the report may state that the patient has epilepsy, Alzheimer's disease, brain tumor, etc. In the medical report, extract the condition status of the patient's brain disease. This status usually includes the severity of the disease, the progress, the current treatment plan, etc.;

[0049] The wearable device types in S1 include electroencephalogram caps, magnetic resonance imaging monitoring devices for the brain, and smart headbands. By synchronizing the data of the wearable device types with the hospital management terminal through a wireless protocol, the obtained brain physiological data is synchronized to the clinical system, providing a data basis for subsequent safety reminders.

[0050] S2. Obtain the patient's brain physiological data in real time through the wearable device, and at the same time monitor the status data of the wearable device at the same time when collecting data. Set the data anomaly threshold for the brain physiological data according to the status data and the brain disease type;

[0051] The steps of S2 are as follows:

[0052] S2.1. Set the acquisition frequency for the wearable device according to the patient's physical information and brain disease type, and then the wearable device regularly acquires the patient's brain physiological data according to the set acquisition frequency;

[0053] Estimate the appropriate acquisition frequency of brain physiological data through the patient's age, gender, weight, height, etc. Set different acquisition frequencies according to the patient's brain disease type (such as epilepsy, brain tumor, stroke, etc.). For example, for epilepsy patients, a higher frequency (such as collecting data once per second) may be required, while for brain tumor patients, a lower frequency may be needed.

[0054] S2.2. Extract the status data of the wearable device, and at the same time extract the status data at the same time according to the time of collecting the brain physiological data;

[0055] According to the set acquisition frequency, the wearable device (such as an electroencephalogram cap, a magnetic resonance imaging monitoring device for the brain, a smart headband) starts to regularly acquire the patient's brain physiological data, and extracts the data related to the status of the wearable device from the sensor data, such as information about the device's temperature, pressure, battery power, etc. Synchronize these status data with the brain physiological data, and extract the status data corresponding to the acquisition time when the device acquires the brain physiological data;

[0056] S2.3. Obtain the standard status data of the patient's wearable device (the standard status data refers to the status data of the wearable device under normal conditions. Usually, these data are obtained without any diseases or interferences), and then set the data anomaly threshold for the brain physiological data by combining the standard status data, the extracted status data, and the type of brain disease;

[0057] The greater the difference value between the extracted status data and the standard status data, the higher the data anomaly threshold;

[0058] The smaller the difference value between the extracted status data and the standard status data, the lower the data anomaly threshold. When the extracted status data is the same as the standard status data, the data anomaly threshold is 0. The specific formula is as follows:

[0059] D = |S zt - S bz |;

[0060] Where D is the difference value, S zt is the currently collected status data, and S bz is the standard status data;

[0061] T yc = k·D;

[0062] Where T yc is the set anomaly threshold, and k is an adjustment factor used to control the sensitivity of the anomaly threshold;

[0063] The greater the difference value, the more abnormal the status of the device, and the higher the set anomaly threshold. The smaller the difference value, the more normal the status of the device, and the lower the anomaly threshold;

[0064] If the currently collected brain physiological data minus the brain physiological data within the normal range is greater than Tyc, it is considered that the data is abnormal;

[0065] During the monitoring process, the doctor can automatically adjust the anomaly threshold according to the patient's real-time condition (such as disease progression).

[0066] S3. Use the AI model to perform anomaly data analysis by combining the historical brain physiological data, the status data, and the data anomaly threshold, and at the same time extract the development trend of the adjacent time periods of the abnormal data, and make normal adjustments to the abnormal data according to the development trend;

[0067] The steps of S3 are as follows:

[0068] S3.1. Establish an AI model based on the deep learning algorithm;

[0069] S3.2. Combine the historical brain physiological data with the status data at the same time and the data anomaly thresholds corresponding to each status data for anomaly data analysis. When the difference between the historical brain physiological data at a certain moment and the historical brain physiological data at the previous moment exceeds the data anomaly threshold, it is determined that the historical brain physiology at this moment is abnormal data. On the contrary, when the difference between the historical brain physiological data at a certain moment and the historical brain physiological data at the previous moment does not exceed the data anomaly threshold, it is determined that the historical brain physiology at this moment is normal data;

[0070] S3.3. Obtain the development trend of the historical brain physiological data, and at the same time extract the development trend of the adjacent time periods of the abnormal data. Calibrate the abnormal data according to the development trend, and then make normal adjustments to the abnormal data after data calibration. The specific steps are as follows:

[0071] Extract the development trend of historical data: Conduct trend analysis based on the historical brain physiological data (such as EEG data). Common trend analysis methods include using techniques such as sliding windows and time series analysis. Then, fit the historical brain physiological data according to the model to obtain the predicted values for a period of time in the future. The formula is as follows:

[0072] X yc (t + 1) = f(X t );

[0073] Among them, f(X t ) is the trend prediction function, predicting the brain physiological data at the future time t + 1, and X yc is the predicted brain physiological data;

[0074] Analysis of the development trend of adjacent time periods of abnormal data: For the data points marked as abnormal, it is necessary to analyze the data trends in their adjacent time periods and perform calibration. Calculate the trend difference between the abnormal data point X t and the data in the adjacent time period. If the abnormal data point deviates greatly from the trend, the calibration formula is as follows:

[0075]

[0076] Among them, is the abnormal data after data calibration, X t is the abnormal data, and ΔX qs is the trend adjustment value;

[0077] Once the abnormal data is calibrated, it can be adjusted to normal data to ensure the coherence and accuracy of the entire data sequence.

[0078] During the data calibration of abnormal data in S3.3, the disease state of brain diseases is monitored. If a disease condition adjustment occurs in the adjacent period of the abnormal data acquisition time, the abnormal data will not be calibrated, and the abnormal data will be fed back to the hospital management terminal for manual adjustment.

[0079] S4. The AI model collects relevant patient data of the same type of brain disease, then combines the collected relevant patient data with the adjusted historical brain physiological data to predict the development data of brain diseases, and generates corresponding predicted brain physiological data according to the predicted brain disease development data;

[0080] The steps of S4 are as follows:

[0081] S4.1. Collect all patient data through the AI model on the hospital management terminal, extract relevant patient data according to different types of brain diseases from the patient data, and then only save the relevant patient data of the same type of brain disease for this patient safety reminder;

[0082] Collect all patient data: Obtain the health data of all patients from the hospital management terminal, including the basic information of the patients, historical medical history, brain physiological data (such as EEG data), disease types, etc.;

[0083] Screen patient data of specific brain disease types: Perform data screening according to the brain disease types, only retain the patient data related to the target disease, and store the screened data in a new dataset for subsequent analysis and prediction.

[0084] S4.2. Combine the relevant patient data with the adjusted historical brain physiological data to predict the development data of brain diseases for the patients of this safety reminder, obtain the corresponding predicted brain disease development data for this patient, and then generate predicted brain physiological data according to the predicted brain disease development data. The specific steps are as follows:

[0085] Combine the adjusted historical brain physiological data for disease prediction: Preprocess and adjust the screened historical brain physiological data, such as smoothing, denoising, trend calibration, etc., to obtain the adjusted brain physiological data. Extract features from the adjusted historical brain physiological data, such as time-domain features (mean, standard deviation, maximum value, minimum value, etc.), frequency-domain features (spectrum analysis), and other features that may be useful for predicting disease development. Based on the historical brain physiological data and data related to the relevant disease types, use the AI model to predict the future development of brain diseases and obtain the data of the future development of the patient's brain diseases;

[0086] Generate predicted brain physiological data based on the prediction results: According to the prediction results of the brain disease development data, use these data to generate corresponding brain physiological data, which can be achieved through inverse models (such as backpropagation neural networks, generative adversarial networks) or direct regression generation methods. Since there may be certain errors in the predicted data, it is necessary to correct the generated predicted brain physiological data. For example, it can be adjusted according to the known physiological data rules to ensure that the predicted data conforms to physiological rationality.

[0087] S5. Set accurate thresholds and effective ranges according to the disease state of the brain disease type, compare the real-time brain physiological data combined with the predicted brain physiological data with the accurate threshold, determine the effective range of the predicted brain physiological data according to the comparison result, and give a safety reminder to the patient based on the predicted brain physiological data.

[0088] The steps of S5 are as follows:

[0089] S5.1. Set accurate thresholds and effective ranges according to the disease state of the brain disease;

[0090] S5.2. Compare the real-time brain physiological data combined with the predicted brain physiological data with the accurate threshold. When the comparison result shows that the deviation between the real-time brain physiological data and the predicted brain physiological data is less than the accurate threshold, extract the predicted brain physiological data within the corresponding range according to the effective range, and give a safety reminder to the patient based on the predicted brain physiological data within the corresponding range;

[0091] S5.3. When the comparison result shows that the deviation between the real-time brain physiological data and the predicted brain physiological data is greater than the accurate threshold, continue to monitor. The specific steps are as follows:

[0092] Set accurate thresholds and effective ranges: According to clinical practice and the disease state of the brain disease, determine an accurate threshold for determining whether the deviation between the real-time brain physiological data and the predicted brain physiological data is within an acceptable range. This threshold is usually determined by a doctor based on factors such as the type of brain disease, the patient's condition, and the fluctuation range of physiological data;

[0093] Define the effective range: Set an effective predicted brain physiological data range according to the disease development stage. The effective range is usually a tolerance range of brain physiological data, and data changes within this range are considered normal or acceptable;

[0094] Comparison of real-time brain physiological data and predicted brain physiological data: Collect the patient's real-time brain physiological data in real time, obtain the predicted brain physiological data according to the patient's historical data and disease development prediction model, and compare the deviation between the real-time data and the predicted data. The formula is as follows:

[0095] ΔY=|Y s(t) - Y c (t);

[0096] Wherein, ΔY is the deviation, Y s (t) is the real-time brain physiological data at time t, Y c (t) is the predicted brain physiological data at time t;

[0097] When ΔY ≤ accurate threshold, it indicates that the deviation between the real-time data and the predicted data is within the acceptable range;

[0098] When ΔY > accurate threshold, it indicates that the deviation between the real-time data and the predicted data exceeds the expected range, and continuous monitoring is required.

[0099] The second object of the present invention is to provide a wearable state monitoring system for clinical research data security equipment based on artificial intelligence, including a method for monitoring the wearable state of clinical research data security equipment based on artificial intelligence as described in any one of the above, including a data management module, a data prediction module, and a prediction comparison module;

[0100] The data management module is used to obtain the patient's physical information and the type of brain disease, and at the same time obtain the type of wearable equipment. At the same time, it monitors the status data of the wearable equipment collected at the same time, and sets the data anomaly threshold for the brain physiological data according to the status data combined with the type of brain disease;

[0101] The data prediction module is used to use the AI model to analyze the abnormal data by combining the historical brain physiological data with the status data and the data anomaly threshold, and at the same time extract the development trend of the adjacent time periods of the abnormal data, and make normal adjustments to the abnormal data according to the development trend. The AI model collects the relevant patient data of the same type of brain disease, and then combines the collected relevant patient data with the adjusted historical brain physiological data to predict the development data of the brain disease, and generates the corresponding predicted brain physiological data according to the predicted brain disease development data;

[0102] The prediction comparison module is used to set the accurate threshold and the effective range according to the condition status of the type of brain disease, compare the real-time brain physiological data combined with the predicted brain physiological data with the accurate threshold, determine the effective range of the predicted brain physiological data according to the comparison result, and give a safety reminder to the patient according to the predicted brain physiological data.

[0103] The foregoing has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and the descriptions in the specification are only preferred examples of the present invention, which are not used to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of the present invention claimed is defined by the appended claims and their equivalents.

Claims

1. A method for monitoring the wearing status of clinical research data security equipment based on artificial intelligence, characterized in that: It includes the following steps: S1. Obtain the patient's physical information and the type of brain disease, and at the same time obtain the type of wearable device; S2. Real-time obtain the patient's brain physiological data through the wearable device, and at the same time monitor the status data of the wearable device at the same time when collecting data. Set the data anomaly threshold for the brain physiological data according to the status data combined with the type of brain disease; S3. Use the AI model to analyze the abnormal data by combining the historical brain physiological data with the status data and the data anomaly threshold, and at the same time extract the development trend of the adjacent time periods of the abnormal data, and make normal adjustments to the abnormal data according to the development trend; S4. The AI model collects the relevant patient data of the same type of brain disease, and then combines the collected relevant patient data with the adjusted historical brain physiological data to predict the development data of the brain disease, and generates the corresponding predicted brain physiological data according to the predicted development data of the brain disease; S5. Set the accurate threshold and effective range according to the condition status of the type of brain disease. Compare the real-time brain physiological data combined with the predicted brain physiological data with the accurate threshold, determine the effective range of the predicted brain physiological data according to the comparison result, and give a safety reminder to the patient according to the predicted brain physiological data.

2. The method for monitoring the wearing state of a clinical research data security device based on artificial intelligence according to claim 1, characterized in that: In S1, the patient's physical information is obtained through the hospital management terminal, and at the same time, the physical information related to the type of brain disease is extracted from the physical information for storage; Obtain the patient's medical examination report through the hospital management terminal, extract the corresponding type of brain disease of the patient and the condition status of the type of brain disease from the medical examination report, and at the same time extract the type of wearable device equipped for the patient from the medical examination report.

3. A method for monitoring the wearing state of a clinical research data security device based on artificial intelligence according to claim 1, characterized in that: In S1, the types of wearable devices include electroencephalogram caps, magnetic resonance imaging monitoring devices for the brain, and smart headbands. Synchronize the data of the type of wearable device with the hospital management terminal through a wireless protocol, so as to synchronize the obtained brain physiological data to the clinical system, providing a data basis for subsequent safety reminders.

4. A method for monitoring the wearing status of a clinical research data security device based on artificial intelligence according to claim 1, characterized in that: The steps of S2 are as follows: S2.

1. Set the acquisition frequency for the wearable device according to the patient's physical information and the type of brain disease, and then the wearable device regularly acquires the patient's brain physiological data according to the set acquisition frequency; S2.

2. Extract the status data of the wearable device, and at the same time extract the status data at the same time according to the time of collecting the brain physiological data; S2.

3. Obtain the standard status data of the patient's wearable device, and then set the data anomaly threshold for the brain physiological data by combining the standard status data and the extracted status data with the type of brain disease; The greater the difference value between the extracted status data and the standard status data, the higher the data anomaly threshold; The smaller the difference value between the extracted status data and the standard status data, the lower the data anomaly threshold. When the extracted status data is the same as the standard status data, the data anomaly threshold is 0.

5. A method for monitoring the wearing state of a clinical research data security device based on artificial intelligence according to claim 1, characterized in that: The steps of S3 are as follows: S3.

1. Establish an AI model based on a deep learning algorithm; S3.

2. Analyze the abnormal data by combining the historical brain physiological data with the status data at the same time and the data abnormality thresholds corresponding to each status data. When the difference between the historical brain physiological data at a certain moment combined with the historical brain physiological data at the previous moment exceeds the data abnormality threshold, it is determined that the historical brain physiology at this moment is abnormal data. On the contrary, when the difference between the historical brain physiological data at a certain moment combined with the historical brain physiological data at the previous moment does not exceed the data abnormality threshold, it is determined that the historical brain physiology at this moment is normal data; S3.

3. Obtain the development trend of the historical brain physiological data, and at the same time extract the development trend of the adjacent periods of the abnormal data. Calibrate the abnormal data according to the development trend, and then make normal adjustments to the abnormal data after data calibration.

6. The method for monitoring the wearing state of a clinical research data security device based on artificial intelligence according to claim 5, wherein: During the process of data calibration of the abnormal data in S3.3, monitor the disease condition of the brain disease. When the disease condition is adjusted in the adjacent period of the abnormal data collection moment, do not perform data calibration on this abnormal data, and feedback the abnormal data to the hospital management terminal for manual adjustment.

7. A method for monitoring the wearing status of a clinical research data security device based on artificial intelligence according to claim 1, characterized in that: The steps of S4 are as follows: S4.

1. Collect all patient data through the AI model in the hospital management terminal, extract relevant patient data according to different brain disease types from the patient data, and then only save the relevant patient data of the same brain disease type for this patient safety reminder; S4.

2. Combine the relevant patient data with the adjusted historical brain physiological data to predict the development data of the brain disease for the patients of this safety reminder, obtain the predicted brain disease development data corresponding to this patient, and then generate predicted brain physiological data according to the predicted brain disease development data.

8. A method for monitoring the wearing status of a clinical research data security device based on artificial intelligence according to claim 1, characterized in that: The steps of S5 are as follows: S5.

1. Set the accurate threshold and effective range according to the disease condition of the brain disease; S5.

2. Compare the real-time brain physiological data combined with the predicted brain physiological data with the accurate threshold. When the comparison result shows that the deviation between the real-time brain physiological data and the predicted brain physiological data is less than the accurate threshold, extract the predicted brain physiological data within the corresponding range according to the effective range, and give a safety reminder to the patient according to the predicted brain physiological data within the corresponding range; S5.

3. When the comparison result shows that the deviation between the real-time brain physiological data and the predicted brain physiological data is greater than the accurate threshold, continue to monitor.

9. An artificial intelligence-based clinical research data security equipment wearing state monitoring system for implementing an artificial intelligence-based clinical research data security equipment wearing state monitoring method according to any one of claims 1-8, characterized in that: It includes a data management module, a data prediction module, and a prediction comparison module; The data management module is used to obtain the patient's physical information and brain disease type, and at the same time obtain the wearable device type. At the same time, monitor the status data of the data collected by the wearable device at the same time, and set the data abnormality threshold for the brain physiological data according to the status data combined with the brain disease type; The data prediction module is used to analyze abnormal data by using an AI model to combine historical brain physiological data with status data and data anomaly thresholds, and at the same time extract the development trends of adjacent periods of abnormal data, and make normal adjustments to the abnormal data according to the development trends. The AI model collects relevant patient data of the same type of brain disease, and then combines the collected relevant patient data with the adjusted historical brain physiological data to predict the development data of the brain disease, and generates corresponding predicted brain physiological data according to the predicted brain disease development data; The prediction comparison module is used to set accurate thresholds and effective ranges according to the disease status of the brain disease type, compare the real-time brain physiological data combined with the predicted brain physiological data with the accurate thresholds, determine the effective range of the predicted brain physiological data according to the comparison results, and give safety reminders to the patients according to the predicted brain physiological data.

Images